W3C

– DRAFT –
ARIA Authoring Practices Task Force

10 September 2024

Attendees

Present
Adam_Page, Daniel, howard-e, jugglinmike, lola, Matt_King
Regrets
-
Chair
-
Scribe
jugglinmike

Meeting minutes

<jongund> This Jon I am going to be a few minutes late

Setup and Review Agenda

Matt_King: Any requests for change to agenda?

Matt_King: Hearing none, we'll move forward with the agenda as it is

Matt_King: Next meeting: September 17

Matt_King: No meeting September 24 due to TPAC

Publication planning

Matt_King: Previously, we had planned to make a request for publication this Thursday

Matt_King: I know Daniel won't be in the office, but will Shawn be available?

Daniel: Yes; I'll be keeping an eye on this

Matt_King: Great. That will be our plan

Matt_King: It's going to be a fairly light publication, I think

Matt_King: Right now, we have two things that are ready to go and I believe two more things that will definitely be ready

Matt_King: Only feature of the five in the current milestone may not be ready in time: the high-contrast work

Daniel: Got it. I'll communicate with Shawn and Rémi

Tabs with action buttons

github: w3c/aria-practices#3071

<jugglinmike> s/Remi/Rémi/

Adam_Page: I made three examples on the one page, and I removed two as we discussed in the previous meeting

Adam_Page: I had put "presentation=role" on the tab wrapper to contain the tab and the button, but since it was just a "div" element, that wasn't necessary. So I removed it

Matt_King: We don't have any documentation in the tables down below about ARIA actions or about the keyboard interaction changes

Matt_King: That's something I can add

Matt_King: Under "keyboard support", currently we have the "tab" key documented. That one would need to be changed

Matt_King: When you press the "tab" key on the selected tab, you would get to the actions menu button...

Matt_King: If you're tabbing backward, that's the only time this distinction matters

Matt_King: But we don't have "shift+tab" documented in this table, so I would have to add a row to document "shift+tab"

Matt_King: None of the others--left, home--they're not affected

Matt_King: Is that it for keyboard documentation?

Matt_King: I don't think we need to document the keyboard interaction for the menu itself. I wonder if I should add a reference to the "menu" page

Adam_Page: That's what I was thinking

Matt_King: Then for "roles, states, and properties", the only new property is "aria-actions". I think that would be the only change to that table

Adam_Page: I think you're right

Matt_King: Do you have any time to add any of that yourself before some time tomorrow?

Adam_Page: I think so. Can you give me a deadline?

howard-e: I will make the pull request on Thursday morning

Matt_King: Then a good deadline for Adam_Page would be 17:00 PT tomorrow

Adam_Page: That works for me

howard-e: Just to confirm: I usually make the pull requests on Thursday by 13:00 ET, which I believe is 10:00 PT

Adam_Page: I have linting errors in the pull request; can we take a look at those?

Adam_Page: It seemed like linting was set up to run on my local environment, so I'm not sure why they passed on my machine but that the CI tests are failing

howard-e: It mentions aria-actions...

Matt_King: Ah. We have a set of lint exceptions for HTML that are related to either bugs in the HTML linter or new things like this

Matt_King: This hasn't been picked up yet--it doesn't know what's valid or not because the spec isn't final

Matt_King: I think we need to modify the .vnurc file

howard-e: Right; the line you would add there is whatever follows the "error:" in the error message itself

howard-e: This is something you can do in your pull request, Adam_Page

howard-e: There's also some spell check errors here. You'd add those to the "cspell.json" file in the root of the project. It's an object; you just add the words to the root of that object

Adam_Page: I should be able to take care of this, no problem. If I have any trouble, I'll ping you folks

Adam_Page: I shared the deployment preview with a colleague. He tried to use the CodePen and found that it was broken. Is that a problem for now?

Matt_King: We could merge with it broken since this is experimental and not many people will see it...

Matt_King: But why isn't it working?

Adam_Page: I think that because I am setting a new precedent with this example--the example uses the code from another example (the menu pattern's JavaScript and CSS)

Matt_King: I think we've often copied them for that reason, even though doing so creates redundancy

Adam_Page: For what it's worth, the code pen works somewhat; the tab functionality is there--it is just rendered incorrectly and it doesn't have any functionality of the menu

Matt_King: While it might seem more efficient to not duplicate the code; I think it makes sense to copy here to keep the example self-contained

Matt_King: We do have some shared JavaScript, but that's mostly related to the template for the page itself rather than the example it's demonstrating

Adam_Page: Understood. I'll go ahead and just copy those files into the directory for the tabs

Matt_King: Great!

Matt_King: Does anybody else have feedback on this pull request?

Matt_King: Remember that it is experimental, so we expect to continue gathering feedback over time

Matt_King: Hearing none, we'll move forward with just the fixes we've identified today

Matt_King: Thank you, Adam_Page

Live regions practice

Matt_King: I'm trying to recruit help for a page about live regions

Matt_King: Simon Pieters originally drafted a whole bunch of content about practice with live regions

Matt_King: ...but it got delayed as we tried to figure out exactly what to recommend

Matt_King: I think this is getting resolved as the working group works on actions

Matt_King: You can see the pull request here; I essentially refactored Simon's pull request without making any substantial changes

github: https://github.com/w3c/aria-practices/pull/2989/files

Matt_King: It's ready to go if someone wants to pick it up and become our resident expert in live regions

Matt_King: You can consider this a "public service announcement"

github: w3c/aria-practices#2989

jongund: It looks like the preview failed the build

Matt_King: There were definitely some linting errors here

Matt_King: Okay, well then it actually needs some clean-up work!

jongund: I can take a look at these errors

howard-e: It might be an issue with the build repository itself. There's a long-standing issue where if that repository has progressed way beyond the current structure, there could be incompatibilities

howard-e: I see here that the work was done in April, so there may have been significant changes to the "build" repository in the time since

Matt_King: Thank you jongund and howard-e for agreeing to help get this cleaned up a bit

ARIA 1.3

Matt_King: This is another public-service announcement

Matt_King: There is more work that needs to be done for the ARIA 1.3 milestone

Matt_King: There are six open issues for providing information related to ARIA 1.3

Matt_King: I was pretty sure that the draft that they want to send to "candidate recommendation" is either imminent or close

Matt_King: ...so that makes the importance of these issues increasing

Matt_King: We have stuff related to "ARIA details" which is widely used by certain apps

Matt_King: There are also the new braille properties--that seems like something that needs particularly strong guidance

Matt_King: The one for "term and definition no longer allowing author naming" should actually be a pretty simple one

Matt_King: There's also one for "role=comment" and "suggestion"

Matt_King: There's another, which is more of a reference for someone working on "details" and "suggestion". It's some resources for those roles

Matt_King: So there are kind of three really "meaty" ones in here

Matt_King: Does anybody want to work on any or all of this?

Matt_King: I will do some more recruiting at TPAC, as well

Matt_King: Those are the things that we have for content that are in our roadmap right now. If anyone gets bandwidth to take them on, then that would be helpful!

MDN Proposal - next steps

github: w3c/aria-practices#3098

lola: I don't know if there's anything else to add here today because Matt_King has responded to the proposal

lola: My previous action item was to write the proposal

lola: I'm still waiting on Ruth's input. She just got back from leave, though, so she needs a little more time

lola: She's on UK time, so this meeting is a bit late for her, but I encouraged her to comment on this issue. Hopefully she'll be able to do that within the next couple of weeks

lola: TPAC is in two weeks, though, so we may revisit this after TPAC

Matt_King: My comment is consider how to streamline the proposal--how we would maintain the APG task force mission

Matt_King: Is there anybody from the MDN stakeholders that will be present at TPAC?

lola: I can find out. I don't think that Ruth will be at TPAC, but perhaps some of her colleagues at MDN will

Matt_King: That could be a good forum for having a deeper discussion about these options and their tradeoffs

lola: Agreed, though I won't be attending TPAC this year

lola: We can try to figure out a separate meeting time with Ruth

lola: My understanding for the next phase is to follow up with Ruth next week and to schedule a meeting with the stakeholders

lola: Should I invite anyone else from this group?

Matt_King: Anyone who is interested. If we're discussing anything related to an API, then it could be useful for howard-e to be present because he designed something similar for the ARIA-AT project

howard-e: Sounds good

lola: I've spoken about this with Boaz and Chris at Bocoup. Chris wants to stay abreast, so howard-e can also help keep Chris in the loop

lola: Is there anyone else who would like to be involved?

Matt_King: If we had an API in place to pull content from the APG and make that available for syndication for use in other places. Is there? anything in W3.org or WAI--would they end up needing to be a stakeholder in what we do there

Daniel: I'm not sure. I can't think of any at the moment, but I need to double-check, so I'd like to stay in the loop about this.

Daniel: I'm in the central Europe time zone

Matt_King: I was wondering if there might be any technical implications with respect to other sites. I assume it's not really different from any other kind of request hitting the website, but I don't know

Matt_King: I suppose if they were going to be automated in some way (to pull in updates, for example), I wonder how that would work

Matt_King: There are no technical details in the proposal that Lola has posted as a GitHub issue, but it is still worth reading for you, Daniel

Matt_King: Thank you very much, lola!

lola: Thanks, Matt_King

Feedback on Tooltip Pattern Language

github: w3c/aria-practices#3109

Matt_King: there's a note in the tooltip pattern about the mouse-dismiss behavior

Matt_King: The reporter has suggested that the way we worded the second note in the tooltip pattern doesn't conform with WCAG

Matt_King: I don't think that was our intent, but it is vague when it says "on mouseout" (not "on mouseout of the tooltip")

Matt_King: I wonder if other people interpret this language in the way that the person who raised this issue interprets it

Matt_King: The language of concern: "If the tooltip is invoked when the trigger element receives focus, then it is dismissed when it no longer has focus (onBlur). If the tooltip is invoked with mouseIn, then it is dismissed with on mouseOut."

Matt_King: "invoked with mouseIn" -- I interpret that to mean that the user hovers over the element that displays the tooltip

Matt_King: "on mouseOut", that sounds like the mouse moving out of the element that triggers the display of the tooltip, but that is not what WCAG requires, right?

Matt_King: WCAG requires that you can move the mouse into the tooltip itself and the tooltip won't disappear

jongund: We should probably write about the "pointer" rather than the "mouse", but that's a separate issue

jongund: But the reporter's issue makes sense to me

jongund: It's doubtful that the triggering element and the tooltip itself will be contiguous, so there may need to be some kind of delay to allow the pointer to travel from the triggering element to the tooltip

Matt_King: It would be helpful to have some wording which captures the precise "dismiss" conditions

Matt_King: I wonder if WCAG already has some wording to this effect

Adam_Page: I don't think it does

Adam_Page: I think the reporter would be satisfied with a clarification that users can mouse into the tooltip content without dismissing the tooltip

Adam_Page: Because I think they're coming from a place of good faith where the tooltip content is adjacent to their trigger

Matt_King: Does anyone want to propose some wording?

Adam_Page: I'd be glad to take this on

Matt_King: Thank you, Adam_Page! If you just want to make a proposal in the issue itself, then we can get aligned with the reporter and with each other, and then that makes it easier for anyone to pick it up as a formal pull request

Matt_King: I've assigned you, Adam_Page

Matt_King: Thanks, everyone! Have a great rest of your day, wherever you are

Minutes manually created (not a transcript), formatted by scribe.perl version 229 (Thu Jul 25 08:38:54 2024 UTC).

Diagnostics

Succeeded: s/Remi/Rémi/

Failed: s/Remi/Rémi/

Succeeded: s/.vnu.rc/.vnurc/

Succeeded: s/there/there?/

Maybe present: jongund

All speakers: Adam_Page, Daniel, howard-e, jongund, lola, Matt_King

Active on IRC: Adam_Page, jongund, jugglinmike